Vagus nerve stimulators 2026 — Nurosym vs Pulsetto vs Truvaga vs Neuvana Xen vs gammaCore vs Sensate vs Apollo Neuro vs Hoolest VeRelief

What it claims

• 'Activates the vagus nerve in 4 minutes', 'increases HRV by 30%', 'reduces cortisol', 'improves sleep and mood'
• 'Non-invasive, drug-free, side-effect free' anxiety and stress relief
• 'Clinically proven' — usually a link to one small pilot study, occasionally to independent RCTs
• 'Replaces medication', 'better than meditation', 'the missing piece of your recovery stack'

What the label is not telling you

• The vagus nerve is not one thing you 'activate'. It is a bidirectional cranial nerve X with roughly 80% afferent fibres (body → brainstem) and 20% efferent (brainstem → viscera), and its accessible cutaneous branches are limited to two territories: the auricular branch (cymba conchae and tragus of the outer ear — the taVNS target) and the cervical branch (running alongside the carotid sheath in the neck — the tcVNS target). Everything else — vibration on the sternum, on the wrist, in a headband, in a chair — reaches the vagus indirectly at best, through baroreceptor / thermoreceptor / mechanoreceptor pathways that eventually feed the nucleus tractus solitarius. Marketing that says 'vagus nerve stimulation' without specifying the branch and the modality is almost always eliding this distinction.
• Transcutaneous auricular VNS (taVNS) is the branch with the strongest evidence base. Fibres from the auricular branch of the vagus reach the cymba conchae and tragus (Peuker & Filler 2002, Clin Anat), and stimulating them at 25 Hz, 200–500 μs pulse width, 0.5–1.5 mA (below discomfort threshold) produces measurable brainstem BOLD activation on fMRI (Frangos 2015, Brain Stimul; Yakunina 2017), modest HRV shifts (Clancy 2014, Brain Stimul), and — in larger trials — signal in treatment-resistant depression (Rong 2016, J Affect Disord; Kong 2018), refractory epilepsy adjunct, and functional GI disorders. This is the branch Nurosym, Parasym, Pulsetto (some SKUs) and Hoolest VeRelief target.
• Transcutaneous cervical VNS (tcVNS) via handheld or neck-cradle devices is the second branch — and the evidence base is narrower. gammaCore Sapphire (FDA-cleared for migraine and cluster headache, prescription-only in most territories) delivers a defined 5 kHz burst waveform to the cervical vagus and has 20+ randomised trials behind it in headache indications. Truvaga is the same core electrical technology from the same company (electroCore) sold direct-to-consumer as a general 'stress and sleep' device without the medical indication and without the accompanying published RCTs on those outcomes. Pulsetto uses a similar neck-cradle form factor with a different waveform and a much smaller published evidence base (mostly company-sponsored pilots).
• 'HRV goes up' is not the same as 'vagal tone is improved'. A 3-minute HRV window after any calming input — device, breath, humming, cold water on the face, lying down — will look better than the anxious 3-minute window before it. That is regression to the mean and acute parasympathetic response, not chronic vagal-tone adaptation. Only a controlled, matched-baseline design over weeks tells you whether a device is doing something a Physiological Sigh (Balban 2023, Cell Reports Medicine) or a paced humming exhale would not have done for free. The device studies that survive that scrutiny are a much smaller subset than the marketing suggests.
• Neuvana Xen delivers electrical stimulation through in-ear headphones that play music. The claim is that the current, delivered via the ear-canal electrodes, stimulates the auricular vagus while the music masks the sensation. The concept is defensible; the published evidence is thin (small company-sponsored pilots, no independent RCTs of scale). Using it as a music-listening habit is fine; treating it as clinical taVNS is a stretch.
• Sensate is not a vagus nerve stimulator. It is a chest-worn infrasonic resonance device that vibrates at low frequencies (16–100 Hz) on the sternum. The proposed mechanism is bone-conducted vibration reaching the vagus via the thoracic branches and via baroreceptor/mechanoreceptor entrainment. That is a plausible parasympathetic input in the same family as humming, chanting, and diaphragmatic breathing — but calling it 'vagus nerve stimulation' overstates the anatomy. The published evidence (Meier 2023, PLOS ONE; Goldstein 2020) shows modest HRV and subjective-stress improvements versus sham in short trials. Real signal, honest mechanism when described accurately, misleading when marketed as taVNS.
• Apollo Neuro is a wrist- or ankle-worn vibrotactile device, not a vagus nerve stimulator at all. The mechanism is soothing haptic vibration patterns proposed to modulate autonomic state through cutaneous mechanoreceptors. Two published RCTs (Siepmann 2021; small n) show HRV shifts. Whether that reaches the vagus through any specific afferent pathway is unproven. Marketing it inside the same category as taVNS or tcVNS is a category error — Apollo Neuro is a wearable haptic co-regulation cue, not a nerve stimulator.
• Hoolest VeRelief is a handheld taVNS device targeting the neck (auriculotemporal + cervical territory) with a defined waveform (Sig1 protocol) and a growing published evidence base (2022–2025) in performance anxiety, athletic recovery and PTSD-adjacent populations. Smaller footprint than gammaCore/Truvaga, more portable than a full ear-clip taVNS system, defensible mechanism. The evidence base is genuine but still narrow — mostly athlete and performance-anxiety cohorts, less on chronic stress or sleep.
• Contraindications are not decorative. No device in this review should be used by anyone with an implanted electronic device (pacemaker, ICD, cochlear implant, deep-brain stimulator, vagus nerve stimulator implant), active carotid atherosclerosis or history of carotid endarterectomy (tcVNS specifically), a history of seizures or epilepsy without neurology supervision, pregnancy (precaution, insufficient safety data), or an active cardiac arrhythmia. The 'non-invasive = automatically safe' framing is marketing, not medicine.
• Nurosym (€499 + optional monthly subscription) — auricular taVNS, ear-clip, medical-grade — Manufactured in the UK by Parasym Ltd, delivers a defined 25 Hz waveform to the tragus/cymba conchae. Sits on the strongest published external evidence base of the direct-to-consumer devices: research use across depression, POTS, long COVID, atrial fibrillation, IBS. Uncomfortable at higher intensities (real electrical current, real sensation), requires 30–60 min sessions to reach the doses used in trials. Genuinely the most defensible clinical taVNS device in the consumer channel today.
• Parasym (identical hardware to Nurosym, sold to clinicians and research labs) — Same device, different branding, sold as a medical device to practitioners rather than consumers. If your practitioner has one, you are using the same technology as Nurosym.
• Pulsetto (€269–€349 + optional app subscription) — neck-cradle tcVNS, consumer — Delivers a low-current waveform to the sides of the neck via a plastic cradle. Published evidence is company-sponsored pilots; independent replication is not yet in the peer-reviewed literature at scale. The neck-cradle form factor makes it appealing for daily use; the evidence base is not equivalent to Nurosym or gammaCore. Reasonable as an experiment; not clinical.
• Truvaga (€299–€499, no subscription) — handheld tcVNS, consumer — Same core electrical technology as gammaCore Sapphire (both from electroCore) but sold direct-to-consumer without the FDA migraine/cluster indication. The RCT evidence base that supports gammaCore's medical clearance is for headache, not for the general 'stress and sleep' claims Truvaga markets. Solid engineering; the direct-to-consumer marketing outruns the DTC evidence.
• gammaCore Sapphire (prescription only in most territories; €500–€800 + monthly refill card) — handheld tcVNS, medical — The only FDA-cleared device in this review, cleared for episodic and chronic cluster headache, migraine prevention and treatment. 20+ randomised trials in headache. Not a lifestyle device — it is a prescription medical intervention for specific neurological indications. If your neurologist has offered it, that is a different conversation from buying a €300 wellness device.
• Sensate (€249–€299 + optional app subscription) — infrasonic chest-worn — Not a nerve stimulator. Bone-conducted low-frequency vibration on the sternum. Real evidence for modest stress and HRV improvements in short trials. Honestly-described as a vibrotactile parasympathetic input; misleadingly-marketed as 'vagus nerve stimulation'. Genuinely useful adjunct for a pre-sleep or post-work de-escalation routine; not a substitute for taVNS if that is what you actually want.
• Apollo Neuro (€349 + optional app subscription) — wrist/ankle vibrotactile — Not a nerve stimulator. Haptic vibration patterns designed to nudge autonomic state through cutaneous input. Two small published RCTs. Real for what it is (a wearable co-regulation cue); a category error when sold as vagus nerve stimulation.
• Hoolest VeRelief Sig1 (€249–€349) — handheld taVNS, performance/athletic — Defined stimulation protocol targeting cervical/auriculotemporal territory. Published evidence base in performance anxiety and athletic recovery, smaller and newer than Nurosym's evidence. Portable, well-engineered, defensible clinical rationale. Strong second choice behind Nurosym for taVNS specifically.

Effect on the nervous system

The nervous system does not need a device to reach ventral vagal tone. It needs the four inputs that reliably reach the nucleus tractus solitarius: a slow exhale longer than the inhale (physiological sigh, box breathing, coherent breathing at 5.5–6 breaths per minute), a hum or chant or gargle that vibrates the vocal folds and stimulates the pharyngeal branch of the vagus, cold on the face (trigeminal-vagal reflex, mammalian dive response), and orienting attention outward to slow, low-contrast visual and auditory input. Each of those is free, unbounded, and has an evidence base at least as robust as any consumer taVNS device on the market. What a device can do that breath cannot is deliver a defined, replicable dose of afferent input at the exact fibres you are targeting, over sessions long enough to accumulate the neural plasticity the trials measure (30–60 minutes/day for 4–8 weeks in most of the published protocols). That is a real thing. But it is not a shortcut around behaviour, and it is not a substitute for the four foundational inputs. In our coaching cohort, the pattern is consistent: people who buy a taVNS device before they have built a daily breath + humming + cold + orienting habit report modest, plateaued benefit and often abandon the device within 90 days. People who add taVNS on top of an established regulation practice — as the fifth input, not the first — report additive and durable gains, especially in refractory HRV, sleep-onset latency, and post-viral autonomic dysfunction (long COVID, POTS). Devices work as an add-on to behaviour, never as a replacement. Anyone who tells you otherwise is selling a device.

Who it might suit

People with a well-established daily regulation practice (breath + humming + cold + orienting) who have plateaued and want an additive fifth input — Nurosym is the most defensible choice; Hoolest VeRelief is the strong second. People with treatment-resistant depression, treatment-resistant IBS, POTS or long COVID autonomic dysfunction, working with a clinician — Nurosym (or a research-grade Parasym via a practitioner) is the most-studied device for these indications. Migraine or cluster headache patients whose neurologist has offered it — gammaCore Sapphire, on prescription. Athletes and high-performers wanting a portable pre-competition down-regulation tool — Hoolest VeRelief Sig1. People who want a pre-sleep or post-work vibrotactile de-escalation ritual and know that is what they are buying — Sensate on the chest, Apollo Neuro on the wrist. Both work as haptic co-regulation cues; neither is a nerve stimulator. People who want to experiment with tcVNS as a consumer without a headache indication and understand the evidence base is thinner — Pulsetto or Truvaga.

Who should skip it

Anyone with a pacemaker, ICD, cochlear implant, deep-brain stimulator, or implanted vagus nerve stimulator — every electrical device in this review is contraindicated. Anyone with a history of seizures or epilepsy without neurology supervision — taVNS and tcVNS both require medical guidance in this population. Anyone with active carotid atherosclerosis or a history of carotid endarterectomy — tcVNS (gammaCore, Truvaga, Pulsetto) is specifically contraindicated. Anyone pregnant or trying to conceive — insufficient safety data across every device; the precautionary answer is no. Anyone with an active cardiac arrhythmia — clear it with cardiology first. Anyone under 18 — safety and efficacy have not been established in paediatric populations for any of these consumer devices. Anyone using the device as a substitute for building a daily regulation practice — you are buying a plateau. Anyone who reads 'FDA-cleared' on a wellness website and assumes it means 'clinically proven for stress and sleep' — gammaCore is FDA-cleared for migraine and cluster headache, not for anxiety, sleep or general wellness. The consumer-channel devices in this review are not FDA-cleared for any indication.

Bottom line

Honest hierarchy for 2026: (1) €0 — build the four foundational inputs first. Physiological sigh, humming or chanting exhale, cold on the face, orienting attention outward. Ten minutes a day, 30 days, before any device purchase. This moves vagal tone more than any device in this review. (2) Nurosym (€499) — the most-studied direct-to-consumer taVNS device, the branch with the strongest evidence base, defensible for depression/IBS/POTS/long-COVID adjunct use under practitioner supervision. Buy only after step 1 is established. (3) Hoolest VeRelief Sig1 (€249–€349) — the strongest performance/athletic taVNS option; smaller footprint, newer but growing evidence base. (4) gammaCore Sapphire — prescription only, for headache indications, on the advice of a neurologist. Not a lifestyle purchase. (5) Truvaga (€299–€499) — same engineering pedigree as gammaCore, consumer marketing outruns the consumer-outcomes evidence. Reasonable experiment; not clinical. (6) Pulsetto (€269–€349) — attractive form factor, thinner independent evidence base than Nurosym or Hoolest. Reasonable experiment. (7) Sensate (€249–€299) — genuinely useful vibrotactile pre-sleep tool when described honestly. Not a vagus nerve stimulator. (8) Apollo Neuro (€349) — wearable co-regulation cue. Not a nerve stimulator. (9) Neuvana Xen — clever concept, thin evidence base; fine as a music-listening habit, weak as clinical taVNS. Skip any of the €0 foundational steps and every device on this list becomes theatre. Build the practice first; add the device only when the practice has plateaued.